Myasthenia gravis (MG) is characterized by weakness and fatiguability of ocular, bulbar, and extremity musculature. MG is an autoimmune disease in which there are antibodies that bind to the acetylcholine receptor and interfere with neuromuscular junction transmission. The estimated prevalence is 20/100,000. There are approximately 60,000 persons with MG in the United States. Prednisone is the most commonly used corticosteroid for MG and produces significant clinical improvement. As a result, steroids are often considered the choice in first line immunosuppressive therapy in MG patients. Many MG patients require long term treatment with prednisone to control the disease. However, corticosteroids can have dose-limiting side effects such as generalized immunosuppression, hyperglycemia, hypertension, myopathy, weight gain, cataracts, and osteoporosis. Methotrexate (MTX) is an analogue of folic acid and is an anti-metabolite and a potent inhibitor of dihydrofolate reductase, which inhibits purine and pyrimidine synthesis. Methotrexate is a potent immunosuppressant agent. It has been used effectively in the treatment of autoimmune diseases, such as rheumatoid arthritis and multiple sclerosis. This study is a randomized, double-blinded, controlled trial of methotrexate versus placebo in MG patients who are considered corticosteroid-dependent. Clinical evaluations will take place monthly for 12 months. MG patients, who are being treated with prednisone and who meet inclusion criteria, will be randomized to receive either methotrexate or placebo. For the first four weeks they will receive methotrexate 10 mg weekly (on Fridays) or placebo. At the beginning of week five, if there are no clinical or laboratory side effects, the methotrexate or placebo dose will be increased to 20 mg per week. The primary measure of efficacy will be the prednisone area under the time/dose curve (AUC), which measures the total prednisone doses in 12 month. lf patients receiving methotrexate are clinically improved so they require less prednisone than those receiving placebo, this will be evidence that methotrexate is an effective treatment for MG. A number of secondary clinical end-points that will assess the potential efficacy of methotrexate will also be measured.